One of the major barriers to at-scale deployment of small cells has been the fact that they were confined to licensed spectrum. In each country, just three or four MNOs own that spectrum, and if they find it hard to make the case for small cells – especially indoors – or have other priorities, the opportunity to bring innovative wireless services to enterprises and cities can be wasted.
When shared spectrum is added to the mix, there is the opportunity for a wider variety of service providers to deploy small cells. These providers may have very different starting points for their business cases, compared to a national MNO. Rather than prioritizing wide area mobile broadband for their airwaves, like an MNO whose model rests on attracting millions of subscribers, a more specialized company might build a localized hotzone of small cells, to support a particular enterprise, city or vertical market. The case could still be positive, even with a relatively small number of users, because of the low cost (no spectrum fees, no requirement to build out on a national scale) and the high value of the enterprise services.
The combination of small cells and shared spectrum will have a disruptive effect on the mobile market, especially indoors. Important developments include the decision by the FCC in the US to allow general (unlicensed) access in the 3.5 GHz CBRS band, appointing Spectrum Access System operators to ensure these signals do not interfere with those of licensed spectrum holders in the band, or federal incumbent users. More internationally, LTE is becoming available in 5 GHz unlicensed spectrum via technologies like LTE-LAA and MulteFire; and these will have 5G migration paths in future (although 5G-Unlicensed is unlikely to be standardized until at least 2020).
Beyond the standards themselves, regulators and vendors are looking at ways to broaden access to spectrum, by allowing sharing, though under more control than in the WiFi bands, where congestion and interference can result in depleted user experience. Spectrum orchestrators, to assign frequencies on an on-demand basis as required by a particular service, are emerging, and are part of a broader shift towards network slicing.
Shared bands will generally be relatively high frequency, because these will have the capacity to make sharing practicable – and in 5G, some of the millimeter wave frequencies are expected to be assigned on a shared or dynamic basis in future. High bands lend themselves to dense deployments of small cells because of their high capacity and limited range, and so SCF is playing a significant role in defining how cells will behave in shared spectrum, and what services they could support. Work items in the technology and the business groups are addressing these issues, and will provide a series of specifications, business cases and guidelines to lower the barriers for would-be deployers.
By 2022, according to forecasts by SCF, unlicensed and shared spectrum could account for 36% of small cell deployments. Some of these roll-outs will be carried out by specialist operators; others by MNOs which can use shared spectrum to reduce the cost of offering services – such as deep indoor connectivity with high quality voice – that have traditionally presented challenging ROI.
Shared spectrum also improves the business case for ubiquitous coverage – whether for sparsely populated areas or wide area IoT applications – because it reduces overall cost, and allows local providers to roll out small sub-nets to fill any gaps left by the MNOs.
The subnet approach is a powerful way to make use of shared spectrum, building a localized network of small cells, often with its own local core to provide full control for the enterprise, city or other anchor tenant. The subnet model can be applied to any enterprise which wants its own IT department, or its favorite integrator, to have control over its network and services. Subnets rely on the main mobile network only for wide area connectivity.
With technologies that require a licensed anchor, like LTE-LAA, the enterprise or MVNO might control the 5 GHz small cells while mobile core services went on in the MNO’s core. With MulteFire, there is the option to have a completely self-contained network. This model gives the opportunity to enterprises, venues, cities and non-mobile service providers to invest in their own local infrastructure, and so control quality and services, while not having to make the huge leap into buying licensed spectrum or building wide area coverage.
The important point for the business case is this is not just about coverage or capacity hotspots any more. The pushing of intelligence, processing and cloud services to the edge of the network is making the small cell into a goldmine by adding many more services to the connectivity. Technologies like ETSI’s Multi-access Edge Computing (MEC), or the Cisco-driven OpenFog (now the basis of an IEEE standard), are ways to distribute cloud services widely and bring them close to the users – supported along with radios on small cells equipped with processors, or in separate mini-servers. The distributed topology reduces latency and backhaul costs and improves responsiveness and QoS for users.
And open source technology will lower the barriers to entry for everyone, and on the hardware side, is first emerging in small cells for shared spectrum, with developments like Facebook’s OpenCellular.
We’ll be sharing more through our release program in the coming months. SCF members can continue to progress these topics at our SCF Americas Plenary in December. Further details and a registration link will be available soon.